June 14, 2016

Elmer 101
and the Small Wonder Labs SW30+ Transceiver

Installment #2 ...

A step-by-step analysis and build-up of the classic 2-watt 30-meter superhet transceiver

 by designer Dave Benson K1SWL of Small Wonder Labs.


We had the official kick-off and overview last time, and everyone now should have their kits and started the assembly process!

In this episode, we'll analyze the VFO section, and recap the "build process" up to the STEP 3:  VFO Assembly, Measurement and Test.  At the end of this stage, you'll be seeing the first signals emerge from your board!

73, George N2APB and Joe N2CX

Podcast ... Click to listen, or right-click to download to your local computer.

Chat Window (during the show) ...

<20:14:20> "Allen - AE7TG": Opened and assembled through step 3

<20:14:56> "Anthony WR3T": Opened, but my workbench is out of commission for the next couple weeks.

<20:15:02> "AllanK - W0NLY": Allan has done all three items requested.

<20:15:38> "W0BM Pat": pat w0bm has it also

<20:17:07> "W0BM Pat": and i also am through step 3.

<20:20:05> "Carl W2PTZ": Cannot do any testing until we get back in our home in 4-6 weeks.

<20:20:06> "AllanK - W0NLY": It wasn't clear in the original manual that the J4 power connector shown on the silk screen on the board has been "replaced" by wires to be soldered to the holes in the board.

<20:21:26> "Allen - AE7TG": I noticed the same as W0NLY, reading the original manual lead to the answer that the J4 was in the 'extra' connector kit of the original SW30+ (.100 spaced header pins worked for me)

<20:24:40> "W0BM Pat": what size is the power connector? 5.5 X 2.5?  The power jack does not connect the barrel of the power plug.  looks like the wiper was bent down a bit in the jack.  Just using wires now for testing.  If i know the size, I can get another jack...

<20:28:46> "Mike WA8BXN": I think it is 5.5x2.5.  most of my gear uses 5.5x2.1 so I replaced what came in the kit with that size

<20:30:02> "George N2APB":  It is a 2.5mm center conductor coaxial power jack.  Since it is a bit less common, we included the mating plug for it in the kit.

<20:42:29> "Matt kk4fem": powerpole adapters for everything!

<20:46:10> "Obe - KC4VZT": Mine was sitting at 2.389 MHz until I spread the windings, and it is now 2.405MHz

<20:46:31> "Larry K3PEG": Best to just tack solder that toroid in place until proper freq is determined.

<20:52:10> "Allen - AE7TG": Just a note: The VFO in the block diagram on today’s whiteboard specifies 3 - 3.05 MHz (probably for a SW40+)

<20:52:35> "George N2APB": great catch, tnx

<20:54:50> "AllanK - W0NLY": A practical question:  How do you know which tab on a pot should be treated as the "bottom" end (usually ground end) so that the pot behaves as normally expected (volume increases as shaft is turned clockwise)?  Looking at J2, how do you know which pot tab goes to pin 1 and which goes to pin 3?

<20:54:59> "Obe - KC4VZT": It was in Dave's manual that's how I decided to change the distribution of windings

<20:59:13> "Harold N0JUG": K7QO redrew the SW-30+ schematic back in the day. I found a copy at www.aa0zz.com. It is a bit easier to read.

<21:00:05> "Larry K3PEG": The White Paper description on theory of operation pertains to the 40 meter xcvr, not the SW-30+.

<21:00:33> "George N2APB": yup.  can tune that too

<21:02:44> "Allen - AE7TG": I'm using the Accuprobe (desinged by N2CX) and kitted and sold by Texas QRP I believe... any idea what voltage would be reasonable with it, using the LO setting on the probe?

<21:03:17> "Matt kk4fem": http://n5ese.com/rfprobe1.htm

<21:03:48> "Allen - AE7TG": Thanks

<21:05:29> "Joe N2CX": The Acccuprobe is calibrated to read RMS voltage while the voltages in the manual are peak-to-peak.  To convert peak to peak to RMS you divide by 2.83.  Alternatively, to convert rms to peak-to-peak you multiply by 2.83.

<21:09:15> "Obe - KC4VZT": Yeah most spectrum analyzers are 50 ohm not 1 meg.

<21:11:31> "Allen - AE7TG": I have an older scope but no experience with it, any recommendations on a text to get started with?

<21:13:39> "Obe - KC4VZT": I use rubbing alcohol which works pretty good at removing solder flux on the board.

<21:15:56> "Matt kk4fem": Allen, check out w2aew's youtube channel, he's got some good introductory videos on oscopes.

<21:16:02> "Larry K3PEG": I use high purity (very low water content) isopropyl alcohol. I heard that drug stores may stock it. Rubbing alcohol leaves behind "white" spots on the pcb.

<21:16:32> "Larry K3PEG": Receive current is 30-40 mA

<21:16:37> "Rick K3IND": Isn't there such a thing as water-based flux?

<21:17:34> "Joe N2CX": There is, but I don't like the water-based stuff.  Also lead free solder is baaaad!

<21:17:34> "Carl-- AA9RF": yes,... Most newer solder does use water based flux

<21:18:23> "Larry K3PEG": You may want to coat the sanded edges of the pretty red kit box with a RED Magic marker pen. Looks neat, IMHO.

<21:18:45> "Joe N2CX": Good tip Larry!

<21:22:27> "Obe - KC4VZT": It was fun and looks pretty when finished.

<21:23:25> "Larry K3PEG": I always scrape away the enamel from a side of the wire before applying solder- always tin up to the edge of the toroid.

<21:24:22> "Joe N2CX": Alas the AADE proprietor is an SK and I don't think anyone else picked it up.

<21:24:43> "Matt kk4fem": They are no longer selling the AADE meter...I found a kit from G4HUP as an alternative

<21:24:52> "Mike WA8BXN": I've been assembling a board using the kit provided parts and posting my comments on the yahoo web page https://groups.yahoo.com/neo/groups/CWTD/info

<21:25:21> "Carl-- AA9RF": Larry Nice thing about the "Solder-Ease" type is NO more scraping  :)  Just pre-tin and go...


<21:26:21> "Carl W2PTZ": Need a good place to get an anti static mat

<21:26:39> "Larry K3PEG": Guys, take a look at this $23 device-->  http://www.banggood.com/BM8910-Handheld-Smart-SMD-Tester-Tweezers-Resistor-Capacitor-Diode-Intelligent-Testing-Clips-p-1049706.html

<21:27:41> "Carl-- AA9RF": AADE L/C Meter MAY,... Still be available from memorsex.com   Milestone Technologies

<21:27:44> "Obe - KC4VZT": Jameco has a 11.25" x 22 " anti-static mat for 8.95



CWTD Episode #79:  Elmer 101 and the SW30+ Transceiver  ...

VFO Assembly & Test


First, a few Notes: 

  1. Kits are sold out ... But some of Craig's nifty red Enclosure Kits are left.  Nice for accessories ;-)

  2. You can get the FreqMite Kits now from 4SQRP

  3. Accessories in progress ... Keyer Kit, Display,  Arduino control, Spectrum output

  4. Aside:  I found a really nice mic preamp/EQ/processor chip for use in a bare D-104 mic ... SSM2166P

  5. For the latest & greatest SW30+ Kit information, see/download the Updated Manual


Let's briefly mention ... the Inventorying and Power Supply steps

The power supply tests are quite basic. They make sure the supply voltage gets connected to the board and the regulator is working. If you don't see near 12 V on the banded end of D13 make sure that you have power connected to the board. Make sure the diode is installed properly and that you have the polarity right from your power supply. Current from the supply to the board should be about 0. A large current draw at this point suggests a solder bridge somewhere (something fairly hard to do on these boards).

C112 Electrolytic capacitor polarity ... The "negative" symbol is negative (duh).  Guess which lead is positive?

Problems with the +8 V measurement could be bad soldering or installing U1 improperly. You should see very close to 8.0 volts more or less independent of the voltage you supply to the board.

Don't proceed to the next step until you get the right measurements. Of course in each of the steps getting the right values does not guarantee that everything is perfect but is probably good enough to proceed.



Block Diagram ... the VFO is at the center of everything!





From the Theory of Operation

The LO uses the Colpitts configuration.  The frequency-determining capacitors are NPO (C0G) monolithic types.  These devices are extremely compact and offer good temperature stability.  C2 and C3 are a voltage divider providing proper injection to U1.  Note: If you want broader frequency coverage, the value of C8 may be increased.  The useful upper limit for C8 is about 1000pF, and especially at the larger values, capacitor (and varicap diode D1) thermal stability will become crucial.  Use NPO/C0G capacitors if possible, if you tinker with this circuit.

The design also uses a varicap tuning diode for tuning coverage.  While a smooth ball-bearing tuning cap and vernier reduction drive are the preferred approach, that choice drives the cost and mechanical complexity of a transceiver up considerably.  The varicap approach supports quite a compact package!  If you’ve just got to have RIT, an outboard circuit that injects an adjustable DC offset into the diode bias network on receive does the trick.


VFO Discussion

This is a big step in the assembly. The basic test is to see if it oscillates. Supply current to the board although not noted should be a few mA. The test point is the base of Q2. Find it on the schematic. The top of R15 is where you can access it easily. Anything we connect there to make some kind of measurement is probably going to have some effect on the circuit.

One easy way to see if its working is with a receiver that tunes to around 2.4 MHz, the frequency that the VFO approximately runs on. You don't need to actually connect the receiver to the board in most cases. If you hear it close to that, great. If not, a commonly reported problem is with L1. Make sure the insulation on the wire was properly removed and its properly soldered in place. If the frequency is off by much (say 100 Khz) make sure you have the right number of turns on the toroid and that all the capacitors are the right value. Note that C7 will be installed later in the assembly process.

A frequency counter could also be used to check to see if its oscillating and near the right frequency. Connecting the counter to the test point will probably change the frequency a bit. Its also possible that if the counter loads down the circuit too much it might not oscillate at all. Simply using a receiver to listen for the signal has some advantages over trying to use a counter!

Once we know we have oscillation on about the right frequency we might want to get an idea if its amplitude is in the ballpark. Using a scope, I measured about half a volt peak to peak at the test point. Again, making that connection does have some impact on the circuit. I also tried using my simple RF probe with a DVM and saw about 4.5 volts there. Note that just trying to use the AC volts function of a DVM won't work very well if at all.

Ultimately, a potentiometer will be connected to J2 (check the schematic). This will allow adjustment of the voltage on D1 to allow tuning. A quick check of the tuning range at this point can be done by simply connecting a jumper between pins 1 and 2 and then between pins 2 and 3. My tuning range turned out to be between 2.4 and 2.43 MHz which is quite close to what it needs to be when we are all done.

Schematic Fragment:  The VFO


Test procedure:

Apply power. Base of Q2 should have around 2.4 MHz sine wave (probe at top of R15).

Measured: 500 mV PP, 2.43 MHz J2 1-2 jumper, 2.40 MHz J2 2-3 jumper

             4.26 V using RF probe into DVM

Disconnect power.


Spectrum View

Now we get to the interesting part, looking at various points in the completed SW30+ with a spectrum analyzer. This should confirm the theoretical discussions and show some possibly unexpected things as well. These tests and images were done by Mike WA8BXN using a Rigol DSA 815.

Here we are looking at the VFO output measured at R15. The highest level signal is at the expected frequency around 2.4 MHz. We also see the second, third and fourth harmonics at decreasing amplitudes. But there is another signal present, noted as Peak 4 in the list.


VFO Component Layout



Troubleshooting Schematic